用来自另一个数据帧特定列的值替换来自数据帧特定列的 Nan 值

Question

I have a problem replacing the nan values in one column of the dataframe with column values of other dataframe. Here's test example:

    Name  Age Name2
0    tom   10   tom
1   nick   15  nick
2   juli   14  juli
3    NaN   12  anne
4  error   17   neo

I want to replace Nan values from column Name (not other columns if there is some Nan values in them) with specific value from other dataframe, for instance Name2 value from this dataframe:

    Name  Age Name2
4  error   17   neo

What I want to get is this:

    Name  Age Name2
0    tom   10   tom
1   nick   15  nick
2   juli   14  juli
3    neo   12  anne
4  error   17   neo

This is test code for this example:

# initialize list of lists 
data = [['tom', 10, 'tom'], ['nick', 15, 'nick'], ['juli', 14, 'juli'], [np.nan, 12, 'anne'], ['error', 17, 'neo']] 

# Create the pandas DataFrame 
df = pd.DataFrame(data, columns = ['Name', 'Age', 'Name2']) 

df1=df[df.Age==np.max(df.Age)]

Then I tried those three methods, but none works - my dataframe always stays with Nan value.

df.Name.fillna(df1.Name2, inplace=True)

df.where(df.Name.isnull(), df1.Name2, axis=0)

df[df.Name.isnull()].Name=df1.Name2

Can you tell me where I'm making mistake?

Answer 1

Here is necessary convert one element Series to scalar:

df.Name.fillna(df1.Name2.iat[0], inplace=True)
#assign output to new DataFrame, test for not missing values
df = df.where(df.Name.notna(), df1.Name2.iat[0])
#use DataFrame.loc for avoid SettingWithCopyWarning
df.loc[df.Name.isnull(), 'Name']=df1.Name2.iat[0]

---

print (df)
    Name  Age Name2
0    tom   10   tom
1   nick   15  nick
2   juli   14  juli
3    neo   12  anne
4  error   17   neo

Detail :

print (df1.Name2)
4    neo
Name: Name2, dtype: object

print (df1.Name2.iat[0])
neo